[ewg] Re: SRP HA dm_multipath testing and questions

Chieng Etta etta at systemfabricworks.com
Mon Apr 16 15:15:11 PDT 2007


Ishai,

Yes. It is rescan time.  At /usr/sbin/srp_daemon.sh, the value of "retries"
is represented rescan time.
I tried it again using default value 300.  After I added new target (issued
"modprobe ib_srp_target" on the 2nd target), the 2nd target was discovered
immediately by initiator but it took almost 5 minutes to be added/attached
into initiator.

Here are the steps I did:
1. Loaded ib_srp_target module on the first target.
2. Started srp_daemon.sh on the initiator (using default rescan time).
   (The first target was added into initiator immediately).
3. Loaded ib_srp_target module on the 2nd target.
   (5 minutes later, the 2nd target was added into initiator).

Thanks,
Etta  

-----Original Message-----
From: Ishai Rabinovitz [mailto:ishai at dev.mellanox.co.il] 
Sent: Sunday, April 15, 2007 2:10 AM
To: Chieng Etta
Cc: 'Moiz Kohari'; 'Scott Weitzenkamp (sweitzen)'; 'Roland Dreier
(rdreier)'; ewg at lists.openfabrics.org; 'Ken L Johnson'; 'openib'
Subject: Re: [ewg] Re: SRP HA dm_multipath testing and questions

Chieng Etta wrote:
> Hi Moiz,
> 
> I tested "adding new storage" on both OFED 1.2-beta1 and OFED 1.2-rc1. 
> I used srp_daemon.sh to discover and add new storage automatically.  
> 
> On OFED 1.2-beta1, the default "retries" value at srp_daemon.sh was set to
> 300 seconds and I changed it to 60 seconds. The initiator discovered the
new
> target right away, but it took a few minutes to add the new target and new
> path.  
> 
> On OFED 1.2-rc1, I changed the "retries" value at srp_daemon.sh to 30
> seconds.  The initiator discovered the new target, added target and added
> path within 30 seconds. 
> 
> Thanks,
> Etta

Etta, 

I guess you mean "Rescan time" and not "retries".

In any case The "Rescan time" should not effect the time it takes to add the
new target. Srp_daemon registers to get traps notices, so it knows when a
new machine joins the fabric and it can check if it is a target. So the
discovery of the new new target should be immediate on any "Rescan time"
value.

The "Rescan time" option is just to be on the safe side in case of rare race
conditions.

If someone finds out a test that causes srp_daemon to consistently discover
a target too late (only on the Rescan) please report it.


Ishai




More information about the ewg mailing list